Awọn idun eto idanilaraya 6 ni iṣẹ Kubernetes [ati ojutu wọn]

Ni awọn ọdun ti lilo Kubernetes ni iṣelọpọ, a ti ṣajọ ọpọlọpọ awọn itan ti o nifẹ si ti bii awọn idun ni ọpọlọpọ awọn paati eto ti o yori si aibikita ati / tabi awọn abajade ti ko ni oye ti o kan iṣẹ awọn apoti ati awọn adarọ-ese. Ninu nkan yii a ti ṣe yiyan ti diẹ ninu awọn ti o wọpọ julọ tabi awọn ti o nifẹ si. Paapaa ti o ko ba ni orire rara lati pade iru awọn ipo bẹ, kika nipa iru awọn itan iwadii kukuru bẹ - paapaa “ọwọ akọkọ” - jẹ igbadun nigbagbogbo, ṣe kii ṣe bẹẹ?…

Itan 1. Supercronic ati Docker ikele

Lori ọkan ninu awọn iṣupọ, a gba lorekore Docker tio tutunini, eyiti o dabaru pẹlu iṣẹ ṣiṣe deede ti iṣupọ naa. Ni akoko kanna, atẹle naa ni a ṣe akiyesi ninu awọn akọọlẹ Docker:

level=error msg="containerd: start init process" error="exit status 2: "runtime/cgo: pthread_create failed: No space left on device
SIGABRT: abort
PC=0x7f31b811a428 m=0

goroutine 0 [idle]:

goroutine 1 [running]:
runtime.systemstack_switch() /usr/local/go/src/runtime/asm_amd64.s:252 fp=0xc420026768 sp=0xc420026760
runtime.main() /usr/local/go/src/runtime/proc.go:127 +0x6c fp=0xc4200267c0 sp=0xc420026768
runtime.goexit() /usr/local/go/src/runtime/asm_amd64.s:2086 +0x1 fp=0xc4200267c8 sp=0xc4200267c0

goroutine 17 [syscall, locked to thread]:
runtime.goexit() /usr/local/go/src/runtime/asm_amd64.s:2086 +0x1

…

Ohun ti o nifẹ si wa julọ nipa aṣiṣe yii ni ifiranṣẹ naa: pthread_create failed: No space left on device. Ikẹkọ ni kiakia iwe salaye pe Docker ko le ṣe orita ilana kan, eyiti o jẹ idi ti o di didi lorekore.

Ni ibojuwo, aworan atẹle ni ibamu si ohun ti n ṣẹlẹ:

Ipo kanna ni a ṣe akiyesi lori awọn apa miiran:

Ni awọn apa kanna a rii:

root@kube-node-1 ~ # ps auxfww | grep curl -c
19782
root@kube-node-1 ~ # ps auxfww | grep curl | head
root     16688  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     17398  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     16852  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      9473  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      4664  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     30571  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     24113  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     16475  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      7176  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      1090  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>

O wa jade pe ihuwasi yii jẹ abajade ti podu ṣiṣẹ pẹlu supercronic (IwUlO Lọ ti a lo lati ṣiṣẹ awọn iṣẹ cron ni awọn adarọ-ese):

 _ docker-containerd-shim 833b60bb9ff4c669bb413b898a5fd142a57a21695e5dc42684235df907825567 /var/run/docker/libcontainerd/833b60bb9ff4c669bb413b898a5fd142a57a21695e5dc42684235df907825567 docker-runc
|   _ /usr/local/bin/supercronic -json /crontabs/cron
|       _ /usr/bin/newrelic-daemon --agent --pidfile /var/run/newrelic-daemon.pid --logfile /dev/stderr --port /run/newrelic.sock --tls --define utilization.detect_aws=true --define utilization.detect_azure=true --define utilization.detect_gcp=true --define utilization.detect_pcf=true --define utilization.detect_docker=true
|       |   _ /usr/bin/newrelic-daemon --agent --pidfile /var/run/newrelic-daemon.pid --logfile /dev/stderr --port /run/newrelic.sock --tls --define utilization.detect_aws=true --define utilization.detect_azure=true --define utilization.detect_gcp=true --define utilization.detect_pcf=true --define utilization.detect_docker=true -no-pidfile
|       _ [newrelic-daemon] <defunct>
|       _ [curl] <defunct>
|       _ [curl] <defunct>
|       _ [curl] <defunct>
…

Iṣoro naa ni eyi: nigbati iṣẹ-ṣiṣe kan ba ṣiṣẹ ni supercronic, ilana naa ti yọ nipasẹ rẹ ko le fopin si tọ, titan sinu Zombie.

Daakọ: Lati jẹ kongẹ diẹ sii, awọn ilana ti wa ni ifasilẹ nipasẹ awọn iṣẹ-ṣiṣe cron, ṣugbọn supercronic kii ṣe eto init ati pe ko le “gba” awọn ilana ti awọn ọmọ rẹ gbe jade. Nigbati awọn ifihan agbara SIGHUP tabi SIGTERM ba dide, wọn ko kọja si awọn ilana ọmọ, eyiti o mu ki awọn ilana ọmọ ko fopin si ati pe o ku ni ipo Zombie. O le ka diẹ sii nipa gbogbo eyi, fun apẹẹrẹ, ni iru ohun article.

Awọn ọna meji lo wa lati yanju awọn iṣoro:

1. Gẹgẹbi adaṣe igba diẹ - mu nọmba awọn PIDs pọ si ninu eto ni aaye kan ni akoko:

          /proc/sys/kernel/pid_max (since Linux 2.5.34)
                 This file specifies the value at which PIDs wrap around (i.e., the value in this file is one greater than the maximum PID).  PIDs greater than this  value  are  not  allo‐
                 cated;  thus, the value in this file also acts as a system-wide limit on the total number of processes and threads.  The default value for this file, 32768, results in the
                 same range of PIDs as on earlier kernels

2. Tabi ṣe ifilọlẹ awọn iṣẹ ṣiṣe ni supercronic kii ṣe taara, ṣugbọn lilo kanna tini, eyi ti o ni anfani lati fopin si awọn ilana ti tọ ati ki o ko spawn Ebora.

Itan 2. "Awọn Ebora" nigba piparẹ ẹgbẹ kan

Kubelet bẹrẹ jijẹ Sipiyu pupọ:

Ko si ẹnikan ti yoo fẹran eyi, nitorinaa a ni ihamọra ara wa lofinda o si bẹrẹ si koju iṣoro naa. Abajade iwadi naa jẹ bi wọnyi:

• Kubelet lo diẹ sii ju idamẹta ti akoko Sipiyu rẹ nfa data iranti lati gbogbo awọn ẹgbẹ:

  

• Ninu atokọ ifiweranṣẹ awọn olupilẹṣẹ kernel o le wa fanfa ti isoro. Ni kukuru, aaye naa wa si eyi: orisirisi tmpfs awọn faili ati awọn miiran iru ohun ti wa ni ko patapata kuro lati awọn eto nigbati piparẹ a cgroup, awọn ti a npe ni memcg zombie. Laipẹ tabi nigbamii wọn yoo paarẹ lati kaṣe oju-iwe, ṣugbọn iranti pupọ wa lori olupin naa ati pe ekuro ko rii aaye ni sisọ akoko lori piparẹ wọn. Ìdí nìyẹn tí wọ́n fi ń kó wọn jọ. Kini idi ti eyi paapaa n ṣẹlẹ? Eyi jẹ olupin pẹlu awọn iṣẹ cron ti o ṣẹda awọn iṣẹ tuntun nigbagbogbo, ati pẹlu wọn awọn adarọ-ese tuntun. Nitorinaa, awọn ẹgbẹ tuntun ni a ṣẹda fun awọn apoti ninu wọn, eyiti yoo paarẹ laipẹ.
• Kilode ti cAdvisor ni kubelet ṣe npadanu akoko pupọ? Eyi rọrun lati rii pẹlu ipaniyan ti o rọrun julọ time cat /sys/fs/cgroup/memory/memory.stat. Ti o ba wa lori ẹrọ ti o ni ilera iṣẹ naa gba awọn aaya 0,01, lẹhinna lori cron02 iṣoro o gba awọn aaya 1,2. Ohun naa ni pe cAdvisor, eyiti o ka data lati sysfs laiyara, gbiyanju lati ṣe akiyesi iranti iranti ti a lo ninu awọn akojọpọ Zombie.
• Lati yọ awọn Ebora kuro ni agbara, a gbiyanju piparẹ awọn caches bi a ti ṣeduro ni LKML: sync; echo 3 > /proc/sys/vm/drop_caches, - ṣugbọn ekuro naa wa ni idiju diẹ sii o si kọlu ọkọ ayọkẹlẹ naa.

Kin ki nse? Atunse iṣoro naa (dá, ati fun apejuwe kan wo tu ifiranṣẹ) imudojuiwọn kernel Linux soke si ẹya 4.16.

Itan 3. Systemd ati awọn oniwe-òke

Lẹẹkansi, kubelet n gba ọpọlọpọ awọn orisun lori diẹ ninu awọn apa, ṣugbọn ni akoko yii o n gba iranti pupọ:

Ó hàn gbangba pé ìṣòro kan wà pẹ̀lú systemd tí a lò nínú Ubuntu 16.04, ó sì máa ń ṣẹlẹ̀ nígbà tí a bá ń ṣàkóso àwọn àsopọ̀ tí a ṣẹ̀dá fún ìsopọ̀ subPath lati ConfigMap's tabi asiri's. Lẹhin ti awọn podu ti pari iṣẹ rẹ awọn systemd iṣẹ ati awọn oniwe-iṣẹ òke wà ninu eto. Ni akoko pupọ, nọmba nla ti wọn kojọpọ. Awọn ọran paapaa wa lori koko yii:

1. #5916;
2. kubernetes # 57345.

Ikẹhin eyiti o tọka si PR ni systemd: #7811 (oro ni systemd- #7798).

Iṣoro naa ko si nibẹ mọ Ubuntu 18.04, ṣugbọn ti o ba fẹ tẹsiwaju lilo Ubuntu 16.04, o le rii pe ojutu wa lori koko yii wulo.

Nitorinaa a ṣe DaemonSet atẹle:

---
apiVersion: extensions/v1beta1
kind: DaemonSet
metadata:
  labels:
    app: systemd-slices-cleaner
  name: systemd-slices-cleaner
  namespace: kube-system
spec:
  updateStrategy:
    type: RollingUpdate
  selector:
    matchLabels:
      app: systemd-slices-cleaner
  template:
    metadata:
      labels:
        app: systemd-slices-cleaner
    spec:
      containers:
      - command:
        - /usr/local/bin/supercronic
        - -json
        - /app/crontab
        Image: private-registry.org/systemd-slices-cleaner/systemd-slices-cleaner:v0.1.0
        imagePullPolicy: Always
        name: systemd-slices-cleaner
        resources: {}
        securityContext:
          privileged: true
        volumeMounts:
        - name: systemd
          mountPath: /run/systemd/private
        - name: docker
          mountPath: /run/docker.sock
        - name: systemd-etc
          mountPath: /etc/systemd
        - name: systemd-run
          mountPath: /run/systemd/system/
        - name: lsb-release
          mountPath: /etc/lsb-release-host
      imagePullSecrets:
      - name: antiopa-registry
      priorityClassName: cluster-low
      tolerations:
      - operator: Exists
      volumes:
      - name: systemd
        hostPath:
          path: /run/systemd/private
      - name: docker
        hostPath:
          path: /run/docker.sock
      - name: systemd-etc
        hostPath:
          path: /etc/systemd
      - name: systemd-run
        hostPath:
          path: /run/systemd/system/
      - name: lsb-release
        hostPath:
          path: /etc/lsb-release

... ati pe o nlo iwe afọwọkọ wọnyi:

#!/bin/bash

# we will work only on xenial
hostrelease="/etc/lsb-release-host"
test -f ${hostrelease} && grep xenial ${hostrelease} > /dev/null || exit 0

# sleeping max 30 minutes to dispense load on kube-nodes
sleep $((RANDOM % 1800))

stoppedCount=0
# counting actual subpath units in systemd
countBefore=$(systemctl list-units | grep subpath | grep "run-" | wc -l)
# let's go check each unit
for unit in $(systemctl list-units | grep subpath | grep "run-" | awk '{print $1}'); do
  # finding description file for unit (to find out docker container, who born this unit)
  DropFile=$(systemctl status ${unit} | grep Drop | awk -F': ' '{print $2}')
  # reading uuid for docker container from description file
  DockerContainerId=$(cat ${DropFile}/50-Description.conf | awk '{print $5}' | cut -d/ -f6)
  # checking container status (running or not)
  checkFlag=$(docker ps | grep -c ${DockerContainerId})
  # if container not running, we will stop unit
  if [[ ${checkFlag} -eq 0 ]]; then
    echo "Stopping unit ${unit}"
    # stoping unit in action
    systemctl stop $unit
    # just counter for logs
    ((stoppedCount++))
    # logging current progress
    echo "Stopped ${stoppedCount} systemd units out of ${countBefore}"
  fi
done

ati pe o nṣiṣẹ ni gbogbo iṣẹju 5 nipa lilo supercronic ti a mẹnuba tẹlẹ. Dockerfile rẹ dabi eyi:

FROM ubuntu:16.04
COPY rootfs /
WORKDIR /app
RUN apt-get update && 
    apt-get upgrade -y && 
    apt-get install -y gnupg curl apt-transport-https software-properties-common wget
RUN add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu xenial stable" && 
    curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add - && 
    apt-get update && 
    apt-get install -y docker-ce=17.03.0*
RUN wget https://github.com/aptible/supercronic/releases/download/v0.1.6/supercronic-linux-amd64 -O 
    /usr/local/bin/supercronic && chmod +x /usr/local/bin/supercronic
ENTRYPOINT ["/bin/bash", "-c", "/usr/local/bin/supercronic -json /app/crontab"]

Itan 4. Idije nigba ṣiṣe eto awọn pods

A ṣe akiyesi pe: ti a ba ni adarọ-ese ti a gbe sori ipade kan ati pe aworan rẹ ti fa jade fun igba pipẹ, lẹhinna podu miiran ti “lu” oju ipade kanna yoo rọrun. ko bẹrẹ lati fa aworan ti awọn titun podu. Dipo, o duro titi aworan ti adarọ ese ti tẹlẹ yoo fa. Bi abajade, adarọ-ese ti o ti ṣeto tẹlẹ ati pe aworan rẹ le ti ṣe igbasilẹ ni iṣẹju kan yoo pari ni ipo ti containerCreating.

Awọn iṣẹlẹ yoo dabi iru eyi:

Normal  Pulling    8m    kubelet, ip-10-241-44-128.ap-northeast-1.compute.internal  pulling image "registry.example.com/infra/openvpn/openvpn:master"

O wa jade pe aworan kan lati iforukọsilẹ ti o lọra le dènà imuṣiṣẹ fun ipade.

Laanu, ko si ọpọlọpọ awọn ọna jade ninu ipo naa:

1. Gbiyanju lati lo iforukọsilẹ Docker rẹ taara ninu iṣupọ tabi taara pẹlu iṣupọ (fun apẹẹrẹ, Iforukọsilẹ GitLab, Nesusi, ati bẹbẹ lọ);
2. Lo awọn ohun elo bii kraken.

Itan 5. Awọn apa duro nitori aini iranti

Lakoko iṣẹ ti awọn ohun elo lọpọlọpọ, a tun pade ipo kan nibiti ipade kan da duro patapata lati wa ni wiwọle: SSH ko dahun, gbogbo awọn daemons ibojuwo ṣubu, ati lẹhinna ko si nkankan (tabi o fẹrẹ jẹ ohunkohun) ailorukọ ninu awọn akọọlẹ.

Emi yoo sọ fun ọ ni awọn aworan ni lilo apẹẹrẹ ti ipade kan nibiti MongoDB ti ṣiṣẹ.

Eyi ni ohun ti oke dabi si ijamba:

Ati bii eyi - после ijamba:

Ni ibojuwo, fo didasilẹ tun wa, nibiti ipade naa dẹkun lati wa:

Nitorinaa, lati awọn sikirinisoti o han gbangba pe:

1. Ramu lori ẹrọ naa sunmọ opin;
2. Fofo didasilẹ wa ni lilo Ramu, lẹhin eyi wiwọle si gbogbo ẹrọ jẹ alaabo airotẹlẹ;
3. Iṣẹ-ṣiṣe nla kan de Mongo, eyiti o fi ipa mu ilana DBMS lati lo iranti diẹ sii ati kika ni agbara lati disk.

Ó hàn gbangba pé tí ó bá wà nínú Linux Ìrántí òfo ti tán (títẹ̀ ìrántí máa ń wáyé) kò sì sí ìyípadà kankan, lẹ́yìn náà si Nigbati apaniyan OOM ba de, iṣe iwọntunwọnsi le dide laarin jiju awọn oju-iwe sinu kaṣe oju-iwe ati kikọ wọn pada si disk. Eyi ni a ṣe nipasẹ kswapd, eyiti o fi igboya sọ ọpọlọpọ awọn oju-iwe iranti bi o ti ṣee ṣe fun pinpin atẹle.

Laanu, pẹlu ẹru I/O nla pọ pẹlu iye kekere ti iranti ọfẹ, kswapd di igo ti gbogbo eto, nítorí wọ́n ti so mọ́ ọn gbogbo awọn ipin (awọn aṣiṣe oju-iwe) ti awọn oju-iwe iranti ninu eto naa. Eyi le tẹsiwaju fun igba pipẹ ti awọn ilana ko ba fẹ lati lo iranti mọ, ṣugbọn o wa titi ni eti pupọ ti abyss apaniyan OOM.

Ibeere adayeba ni: kilode ti apaniyan OOM ṣe pẹ to bẹ? Ninu aṣetunṣe lọwọlọwọ, apani OOM jẹ aṣiwere pupọ: yoo pa ilana naa nikan nigbati igbiyanju lati pin oju-iwe iranti ba kuna, ie. ti aṣiṣe oju-iwe ba kuna. Eyi ko ṣẹlẹ fun igba pipẹ, nitori kswapd fi igboya ṣe awọn oju-iwe iranti laaye, sisọ kaṣe oju-iwe (gbogbo disk I / O ninu eto, ni otitọ) pada si disk. Ni awọn alaye diẹ sii, pẹlu apejuwe awọn igbesẹ ti o nilo lati yọkuro iru awọn iṣoro ninu ekuro, o le ka nibi.

Iwa yii yẹ ki o mu dara pẹlu mojuto kan Linux 4.6 +.

Itan 6. Pods di ni isunmọtosi ni ipinle

Ni diẹ ninu awọn iṣupọ, ninu eyiti ọpọlọpọ awọn adarọ-ese ti n ṣiṣẹ gaan, a bẹrẹ lati ṣe akiyesi pe pupọ julọ wọn “duro” fun igba pipẹ ni ipinlẹ naa. Pending, botilẹjẹpe awọn apoti Docker funrararẹ ti nṣiṣẹ tẹlẹ lori awọn apa ati pe o le ṣiṣẹ pẹlu ọwọ.

Pẹlupẹlu, ni describe ko si ohun ti ko tọ:

  Type    Reason                  Age                From                     Message
  ----    ------                  ----               ----                     -------
  Normal  Scheduled               1m                 default-scheduler        Successfully assigned sphinx-0 to ss-dev-kub07
  Normal  SuccessfulAttachVolume  1m                 attachdetach-controller  AttachVolume.Attach succeeded for volume "pvc-6aaad34f-ad10-11e8-a44c-52540035a73b"
  Normal  SuccessfulMountVolume   1m                 kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "sphinx-config"
  Normal  SuccessfulMountVolume   1m                 kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "default-token-fzcsf"
  Normal  SuccessfulMountVolume   49s (x2 over 51s)  kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "pvc-6aaad34f-ad10-11e8-a44c-52540035a73b"
  Normal  Pulled                  43s                kubelet, ss-dev-kub07    Container image "registry.example.com/infra/sphinx-exporter/sphinx-indexer:v1" already present on machine
  Normal  Created                 43s                kubelet, ss-dev-kub07    Created container
  Normal  Started                 43s                kubelet, ss-dev-kub07    Started container
  Normal  Pulled                  43s                kubelet, ss-dev-kub07    Container image "registry.example.com/infra/sphinx/sphinx:v1" already present on machine
  Normal  Created                 42s                kubelet, ss-dev-kub07    Created container
  Normal  Started                 42s                kubelet, ss-dev-kub07    Started container

Lẹhin ti n walẹ diẹ, a ṣe arosinu pe kubelet nìkan ko ni akoko lati firanṣẹ gbogbo alaye nipa ipo ti awọn adarọ-ese ati awọn idanwo igbesi aye / imurasilẹ si olupin API.

Ati lẹhin ikẹkọ iranlọwọ, a rii awọn aye wọnyi:

--kube-api-qps - QPS to use while talking with kubernetes apiserver (default 5)
--kube-api-burst  - Burst to use while talking with kubernetes apiserver (default 10) 
--event-qps - If > 0, limit event creations per second to this value. If 0, unlimited. (default 5)
--event-burst - Maximum size of a bursty event records, temporarily allows event records to burst to this number, while still not exceeding event-qps. Only used if --event-qps > 0 (default 10) 
--registry-qps - If > 0, limit registry pull QPS to this value.
--registry-burst - Maximum size of bursty pulls, temporarily allows pulls to burst to this number, while still not exceeding registry-qps. Only used if --registry-qps > 0 (default 10)

Bi a ti ri, awọn iye aiyipada jẹ ohun kekere, ati ni 90% wọn bo gbogbo awọn aini ... Sibẹsibẹ, ninu ọran wa eyi ko to. Nitorinaa, a ṣeto awọn iye wọnyi:

--event-qps=30 --event-burst=40 --kube-api-burst=40 --kube-api-qps=30 --registry-qps=30 --registry-burst=40

ati tun bẹrẹ awọn kubelets, lẹhin eyi a rii aworan atẹle ni awọn aworan ti awọn ipe si olupin API:

... ati bẹẹni, ohun gbogbo bẹrẹ lati fo!

PS

Fun iranlọwọ wọn ni gbigba awọn idun ati ngbaradi nkan yii, Mo ṣe afihan ọpẹ nla mi si ọpọlọpọ awọn ẹlẹrọ ti ile-iṣẹ wa, ati ni pataki si ẹlẹgbẹ mi lati ẹgbẹ R&D wa Andrey Klimentyev (zuzzas).

PPS

Ka tun lori bulọọgi wa:

• «kubectl-debug itanna fun yokokoro ni Kubernetes pods».
• Awọn imọran Kubernetes & ẹtan loop:
  • «Gbigbe awọn orisun nṣiṣẹ ni iṣupọ kan si iṣakoso Helm 2»;
  • «Nipa ipin ipade ati awọn ẹru lori ohun elo wẹẹbu kan»;
  • «Wiwọle si awọn aaye dev»;
  • «Iyara soke bootstrap fun tobi infomesonu».

orisun: www.habr.com